Decomposing Tasks
Productivitydecomposing-tasksDecomposing Tasks is an Agent Skill for productivity published in the majiayu000/claude-skill-registry repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.
Break down large tasks into smaller, actionable items. Use when planning sprints, estimating work, or creating implementation plans. Covers task breakdown strategies.
QUICK START
How to use this skill
Bring this guide into your coding agent with a prompt tailored to the tool you use.
- Open your project in Codex.
- Copy and paste the prompt below.
- Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/majiayu000/claude-skill-registry/blob/HEAD/skills/planning/decomposing-tasks/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/decomposing-tasks, but wait for my approval before writing files or running scripts.
Copying this prompt does not install or run the skill. Review third-party files before use.
Decomposing Tasks
Workflows
- Understand Scope: What is the full scope of work?
- Identify Components: What are the major pieces?
- Break Down: Split into 1-2 day tasks
- Order: Define dependencies and sequence
- Validate: Can each task be demo'd/tested?
Task Size Guidelines
Too Big
- "Build the authentication system"
- Cannot be completed in 1-2 days
- Cannot be tested independently
Just Right
- "Create login API endpoint"
- "Add password validation"
- "Implement JWT token generation"
Too Small
- "Add semicolon"
- "Rename variable"
- Not worth tracking separately
Decomposition Strategies
Vertical Slicing
Cut through all layers for one feature.
Feature: User Login
├── API endpoint (POST /login)
├── Input validation
├── Database query
├── JWT generation
├── Error handling
└── Integration test
Horizontal Slicing
Complete one layer at a time.
Database Layer
├── Users table
├── Sessions table
└── Indexes
API Layer
├── Login endpoint
├── Logout endpoint
└── Refresh endpoint
SPIDR Method
- Spike: Research/investigation task
- Path: Main flow implementation
- Interface: API/UI contract
- Data: Storage and schemas
- Rules: Business logic and validation
Task Template
## Task: [Title]
**Description**: [What needs to be done]
**Acceptance Criteria**:
- [ ] [Criterion 1]
- [ ] [Criterion 2]
**Dependencies**: [What must be done first]
**Estimate**: [X hours/points]
Definition of Done
Every task should have:
- Code complete and reviewed
- Tests written and passing
- Documentation updated
- Deployed to staging